Despite the importance of the indigenous microflora of the intestine to human health and disease, the mechanisms which control its composition are unknown, because (a) experimental models used in the past were inadequate, and (b) the flora is controlled by the interplay of numerous control mechanisms. Consequently, studies of single mechanisms cannot yield an understanding of the whole. During the current reporting period we have overcome these shortcomings (a) by using continuous flow (CF) cultures of intestinal flora from conventional mice and (b) by constructing mathematical models of the interplay among ecologic control mechanisms. This approach will continue during the proposed studies. During the current reporting period we have confirmed a hypothesis formulated at the outset of this project, namely that adhesion of indigenous E. coli is not necessary for colonization of the cecum, but is required to maintain the stability of the flora (i.e. to prevent the colonization of invading microorganisms). Surprisingly, the adhesion that was effective for E. coli in the cecum was of the non-specific type, analogous to the association of polystyrene microspheres with the mucus gel. It became also clear during the present study that bacterial populations of the small intestine may contribute to, and may even control bacterial populations in the large intestine. For this reason we plan to study the bacterial ecology of the entire gastrointestinal tract during the proposed renewal period. To accomplish the above described goal we will perfect a mathematical model of the small intestine, based on tubular flow kinetics, that is now in the preliminary stage. This model will be combined with our present type mathematical models which can adequately describe the stomach and cecum, respectively. This arrangement will be followed by a second tubular flow model to represent the colon. The entire sequence will then simulate the gastrointestinal tract. Simultaneously, the present CF culture model will be expanded to 3 consecutive CF cultures of which the effluent of the preceding culture constitutes the inflow to the next one. The 3 CF cultures thus represent the stomach, small intestine and cecum, respectively. Extensive preliminary tests and modifications will likely be necessary until this system is able to duplicate at least the major bacterial interactions that are observed in vivo. Thereafter it will be used to investigate the physiological basis by which some bacterial populations control the growth and colonization of others (e.g. by metabolic competition, production of inhibitors, etc.). Throughout the study we will maintain a continuous interplay between experimental and mathematical approaches, in which the mathematical model decides between alternative hypothesis and suggests new ones, which are then tested in the CF culture model and in gnotobiotic mice.